IBM TranslationManager Version 2.5 - 32bit
Beta - Driver
Enhancements
Users Guide
SUBJECT TO ANY STATUTORY WARRANTIES WHICH CAN NOT BE EXCLUDED, IBM MAKES NO WARRANTIES OR CONDITIONS EITHER EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, THE WARRANTY OF NON-INFRINGEMENT AND THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, REGARDING THE PROGRAM. THE EXCLUSION ALSO APPLIES TO ANY OF IBM'S SUBCONTRACTORS, SUPPLIERS OR PROGRAM DEVELOPERS (COLLECTIVELY CALLED "SUPPLIERS").
NEITHER IBM NOR ITS SUPPLIERS WILL BE LIABLE FOR ANY DIRECT OR INDIRECT DAMAGES, INCLUDING WITHOUT LIMITATION, LOST PROFITS, LOST SAVINGS, OR ANY INCIDENTAL, SPECIAL, OR OTHER ECONOMIC CONSEQUENTIAL DAMAGES, EVEN IF IBM IS INFORMED OF THEIR POSSIBILITY. SOME JURISDICTIONS DO NOT ALLOW THE EXCLUSION OR LIMITATION OF INCIDENTAL OR CONSEQUENTIAL DAMAGES, SO THE ABOVE EXCLUSION OR LIMITATION MAY NOT APPLY TO YOU.
Using several Translation Memory databases for search for proposals gives the user more flexibility. It is no longer necessary that the user merges several Translation Memory databases into one to get the advantage of all of them for one file. Thus it is possible to use e.g a shared LAN Translation Memory for read-only, and a personal Translation Memory for read and write. In this way you can avoid having raw, not proofread translations mixed up with approved translation.
The dialog of the New Folder dialog is changed, so that two listboxes are added which allow to select additional Translation Memory databases for read-only purposes. At most 4 read-only Translation Memory databases are allowed. Also in the folder properties dialog, the two listboxes for Translation Memory selection will appear. The selected read-only TM's are a folder property and inherited to the document, even if this document has a different document Translation Memory attached. If a Translation Memory is already selected as read/write TM, the read-only one will be ignored.
Furthermore, during folder export, if the option: With Translation Memory is marked, an additional option can be checked. Here the user can determine, whether the read-only TM's should be also exported with the folder. In folder import, read-only Translation Memory databases are imported, too.
In the document properties, it is not possible to select read- only Translation Memory databases. This handling is consistent to the fact, that dictionaries can also only selected on folder level. Also the possibility to select read-only Translation Memory databases on document level would cause too much confusion. Thus the document properties dialog is unchanged.
The batch folder import, export, and create is enhanced so that a list of Translation Memory databases can be handled. The /MEM= can now be followed by a list of memory names, as it is already with the dictionary names (/DICT=). The first specified Translation Memory is the Read/Write Translation Memory, the others are the Read-only Translation Memory databases. Whether all databases should be exported with the folder, can be specified via the /OPTIONS= parameter. In addition to MEM and DICT the new option is ROMEM (for read-only Translation Memory).
Currently, it is not possible to edit in the translation in the Translation Environment while the Find and Change dialog is open. However, very often there is a need to edit the find text and then perform the next Find. Therefore the dialog is changed to be a modeless dialog, which can stay open while the user edits in the translation window. Thus this enhancement allows for a more convenient translation and post editing work.
The dialog procedure of the Find and Change dialog is changed, so that the modeless dialog can be supported. Indeed the Find and Change dialog can be moved to any place on the desktop.
The Find and Change dialog corresponds to the window which was active when the dialog has been opened. Thus the dialog can belong to the Translation window, the Other document window or the Original window. Always, when the Find and Change dialog is closed, the focus returns to the window in which the search was done.
In the Translation Memory Edit, the Find and Change dialog remained modal.
Only one Find and Change dialog can be open at a time. When the Translation Environment window is minimized, the Find and Change dialog will be hidden. During the restore or maximize the Find and Change dialog will be activated again.
The modeless Find and Change dialog can be activated either with the click of the mouse or via Alt-F6.
This user-selectable option allows the user to get all hits from all selected dictionaries.
When copying a dictionary proposal into the Editor, sometimes, especially when it is the first word of a sentence, it would be nice to get an upper case first letter.
Description
There are two possible ways for implementing:
Both versions have their advantage/disadvantage. In the second case there is more flexibility. So this approach is implemented.
To invoke the function select Keys... from the Options pull-down and define a key stroke for the function Toggle first character.
This function will not apply for DBCS characters.
When importing a folder which has the same name as a folder, that is already on your machine, it is sometimes necessary to rename the incoming folder to avoid conflicts.
A new field in Folder Import dialog for folder name and is available now. Also the New Dictionary and New Translation Memory dialogs are invoked for the dictionaries and Translation Memory databases imported together with the folder, if they are not already there.
This approach addresses the dictionary and Translation Memory location (target drive + local/shared) problem during folder import.
To avoid problems during later reimport (Scenario: TC sends folder to freelance translator; freelance translator imports it under new name; translates it; the export is now under the new name, too; return to the TC; how should the TC now know what's the real name?) the original name will be stored in the folder properties, so one can always find out about the 'previous real' name.
For unattended TranslationManager maintenance, the Translation Memory Organize function must be invocable via the command line. This would allow a regular organize of some or all Translation Memory databases during off-shift times or over night.
The EQFCMD now has a new task implemented which allows for Translation Memory organize. The syntax is:
EQFCMD /TA=MEMORG /MEM=SAMPLE1
For unattended TranslationManager maintenance, the Translation Memory Import function for documents and folders is invocable via the command line. The properties, which Editor and which markup shall be used for the folder or the individual documents should be definable.
The EQFCMD now has two new option implemented which allow to specify the Editor and the markup to folders and documents. The syntax is:
EQFCMD /TA=DOCIMP ... /EDit=editor /MArkup=markup
or
EQFCMD /TA=FLDIMP ... /EDit=editor /MArkup=markup
When finished a translation project, it is necessary to archive the created Translation Memory databases for the next usage. Here only the really used segments in the Translation Memory databases are needed. A should removed all segments, which are not used in the current project.
Via the Utilities pull-down there is the Archive Translation Memory. This function asks for a folder or a set of documents and fills a specified Translation Memory with just those segments, which are found translated in the folder or set of documents.
When importing a folder which already exists on the machine, the newer document are replacing the older once. This means that either a document is imported and replaces the existing one or it is skipped and the existing one survives. An option to influence this automatic decision on document level is required.
A new Set Document Processing Options dialog is opened, when you try to import a folder, which already exists. This dialog shows all participating documents, which ones are older/newer and what default action (replace or skip) would happen. By selecting some of them you can influence this action. After checking and evtl. changing the action, you start the real import process.
When Using a Translation Memory coming from BOOKMASTER translations for an MRI translation, all trailing line-feeds in the target segments are disturbing the MRI format. An adoption of trailing line-feeds and blanks to the new sources (which means deletion or insertion of line-feeds and blanks) is required.
This function is now there. When ever a proposal is found which is exact, this trimming happens, so that the proposed target sentence has then as many or as few trailing line-feeds or blanks as the current source segment.
Often you have situations where more than one translation is available for a particular source sentence (segment), but for a human it is easy to determine the right one, because you are currently dealing with the same document. This is particular the case if you are dealing with MRI (e.g. panels) and considering that you quite often get updates. The Use latest exact match causes in such situations more harm than they are helpful. This costs a lot of possible performance in terms of throughput, because a human translator has to go to each of those segments and take the right translation. This new implementation will make the right decision and take the correct translation if there is exactly one available from a document with the same name.
A new option is included in the Text Analysis dialog to allow for substitution of an exact match in the case this is the only one available with the same document name.
The option Use exact context match if more than one found has precedence over the setting of Use latest exact match if more than one found.
Furthermore, the same option is added to the Profile Settings dialog in the Translation Environment. The setting of this option is taken into consideration during the automatic substitution.
During batch folder or document analysis this option can be activated via:
/OPtions=ADDCONTEXT
This option automatically switches the ADDAUTO on, too.
The new Word Count utility for OS/2 and Windows gives users the possibility to collect information about the effort, state and history of the translations. The purpose is to make the process of translation steps transparent without valuing about it. The results of the Word Count utility can be displayed in several reports, and used as a qualified base for calculating the translation cost.
A tracking mechanism collects information during several events in the TranslationManager. These events, called process tasks, are the following:
The records containing the information and results of the tracking depend on the process task, and are stored in a logging file in compressed form. So the history of what happened to a certain document can be recovered later on again. There is one logging file per folder. The so-called "history log file" is located in the PROPERTY directory of the folder and is currently named HISTLOG.DAT. New records are added at the end of the history log file. Process tasks and handling in the history log file
The tracking during editing and analysis is done for:
All tracking results are kept in the segmented target file and are lost as soon as the source file is analyzed again. Thus each time when a document is saved in the editor, this tracking information is kept in a record added to the history log file. Not the information for each segment is stored, only the information summed up and classified by the classes above. There is a fixed order in the different types of proposals indicating the best proposal. An exact proposal is the best, a replace proposal is better than a fuzzy proposal, a fuzzy proposal is better than a machine translation (which may be generated by ITM). A replace fuzzy proposal is better than a fuzzy proposal. If more than one proposal exists, only the counting results of the best proposal are summarized to the appropriate class, independent of the selection made by the user. Only those proposals apply that exist at the point in time the segment was translated for the first time. An exact proposal is either summarized to the class Automatic substitution during analysis, or Automatic substitution in the editor or to the class Exact proposal. The number of source words is counted when a segment is activated, i.e. not yet translated segments always have null source words. The number of target words is updated each time a segment is saved in the editor, as well as when the segment is auto- substituted. For the current segment, the number of target words is null if it has not been translated previously, otherwise it is left unchanged. When the source segment is copied into the target document, the number of target words is set to null. For both source and target segment counting, the following rules apply:
The number of modified words, i.e. the number of words differing in the current source and the source of proposal, is counted using a LCS algorithm. Therefore the best proposal is the relevant one. The modified words are saved when the segment is saved the first time. It is not saved when the segment is activated, but not translated. It is not changed when the segment is retranslated. The information on which proposal has actually been copied is not yet saved in the history log file at this point of time. The current segment is counted in the sum of not translated segments. If a segment has been joined, it is counted only as one segment.
The results of the Word Count utility can be displayed in several reports, and used as a qualified base for calculating the translation cost. The reports can be displayed on screen or stored in a file. Each report is attached to the specified folder and can be created for the whole folder or selected documents in the folder.
Currently there are two reports :
It creates a history about the process tasks according to the selected folder or documents. For this report there are three options you can select:
The report only displays the fixed part of the records such as
1 EMPEREOR.TXT 01/10/96 14:10:56 - Document import
If option Detail is selected, the information of the variable parts of contained process tasks are displayed.
1 EMPEREOR.TXT 01/10/96 14:10:56 - Document import ------------------------------------------------------------------------ Format : External Import path : E:\DEMO\EMPEREOR.TXT Document : EMPEREOR.TXT Markup : EQFASCII Translation memory : DEMO Source language : English(U.S.) Target language : German(national) Source replaced : No Target replaced : No
For each document it creates a summary of the word count results about the events Save in the editor, Automatic substitution during analysis and Replace of target document during import. Only these events are tracked by word count information. For this report there are two options you can select:
The following rules describes about the eight class how the word count results are summarized:
If option With totals is selected, a summary of the classes except Sums of untranslated segments, classified by the three classes small, medium and complex segments, is added to each document. In addition a summary of all documents is created.
Every report can be called from the folder list or document list. If you call a report for a selected folder the following is taken into consideration:
The support for long file names allows to process documents having long file names within TranslationManager. Long file names are all file names not following the 8.3 naming scheme as imposed by DOS. Long file names may have a length of up to 256 characters and are not restricted to the 8.3 character set. Long file names may contain whitespace characters and mixed-case characters.
Within TranslationManager the long file names are managed to short file names following the DOS 8.3 naming scheme. The document names displayed to the user are still the long file names. This ensures compatibility to older TranslationManager versions and allows to work with long document names within TranslationManager even if the TranslationManager drive does not support long file names.
If a document with a long file name is exported to a drive not supporting long file names the long file name is preserved in the .LONGNAME extended attribute of the exported file. If such a file is copied to a HPFS drive using the OS/2 workplace shell, the original long name will be restored.
For the user interface there are no significant modifications for the long file name support. All document names displayed to the user are long file names and the column widths and width of entry fields for document names have been enlarged accordingly.
New API functions (EQFCMD) are provided for the delete of folders and documents in batch mode, a ality is added to export of documents (SNOMATCH) and the dictionary export is implemented in batch mode.
The syntax is:
EQFCMD /TA[SK]=FOLDEL /FLD=folder [/QUIET]
where the
[/QUIET]parameter means
[/QUIT] : Do not pop up any message box, but pass messages to caller
via DossExit function.
The syntax is:
EQFCMD /TA[SK]=DOCDEL /FLD=folder /FI[LES]=documents [/QUIET]
where the
[/QUIT]parameter means
[/QUIT] : Do not pop up any message box, but pass messages to caller
via DossExit function.
The syntax is:
EQFCMD /TA[SK]=DOCEXP /FLD=folder /FI[LES]=documents /OP[TIONS]=option [/QUIET]
where the parameters mean
/OP[TIONS]=option : Selects the type of export. The following values
can be specified:
SOURCE export source documents,
TARGET export target document,
SNOMATCH export documents containing
non-matching segments
[/QUIT] : Do not pop up any message box, but pass messages to caller
via DossExit function.
The syntax is:
EQFCMD /TA[SK]=DICEXP /DICT=dictionary /OUT=outname /OV[ERWRITE]=YESNO [/QUIET]
where the parameters mean
[/OV] : if YES is specified an existing output file will be
overwritten,
if NO is specified the export will stop
if the output file exists already
[/QUIT] : Do not pop up any message box, but pass messages to caller
via DossExit function.
The new alias function allows to specify an alias name for documents which will be used for the access to the Translation Memory instead of the actual document name. The alias feature is useful if an update of the document has a different name (e.g. because the version number is included in the document name). Without the alias feature there would be no exact matches from the Translation Memory as the document name are not the same. Using the old document name as alias for the updated document the Translation Memory will return exact matches again. The same alias name may also be specified for a group of documents if the documents are all for the same subject (e.g. the manuals for different printer models). The Translation Memory will return exact matches even if the segment was translated in a different document of the group. The alias name can be specified during document import (external format and batch mode) and in the document properties window. The alias name is displayed in a new column of the document list window. There is no restriction on the syntax of the alias name. As the alias name is not used for access to files it has not to follow the rules for file names (character set and length). The alias name may have up to 256 characters and does not have to be unique.
The alias name for a document is specified in the alias entry field of the document properties window or it can be specified during document import in GUI mode using the document settings dialog or it is specified as parameter of batch mode document import. The parameter for the alias in batch mode is "/ALIAS=alias".
EQFCMD /TASK=DOCIMP /FLD=TEST /FILES=NEW.DOC /ALIAS=OLD.DOCThe alias names are listed in a new column of the document list window. For all access to Translation Memory databases the alias names are used instead of the actual document names. The alias name can be a long file name with up to 256 characters. Actually, as the alias name is not used for any access to disk files, it may contain any characters even if these characters are not valid for file names (e.g. DBCS characters or national characters). TranslationManager does no checks on the entered file names. It is possible to use the same alias for several documents; i.e. the alias name does not have to be unique.
The Translation Memory Match Counting function allows to count the number and type of matches which can be achieved using a specific Translation Memory on a specific document, a group of documents or all documents of a folder. Unlike "Automatic substitution" the does not alter the document(s) nor the Translation Memory and supplies also information about fuzzy matches.
For the Translation memory match information either the "Substitute exact matches in the document automatically" option or the "Count Translation Memory match information" option must be selected.
The syntax is:
EQFCMD /TA[SK]=ANALYZE /FLD=folder
[/FI[LES]=documents]
[/OP[TION]=options]
[/MEM=memory]
[/OV[ERWRITE]=YES/NO]
[/QUIET]
Where the option means:
options = ADDTOMEM,
AUTO,
UNTRANSLATED,
AUTOLAST,
AUTOJOIN,
AUTOCONTEXT,
TMMATCH
The option "Translation Memory match information" has to be selected to display the Translation Memory match data.
Using the batch mode the word count command is the following:
EQFCMD /TA[SK]=WORDCNT
/FLD=folder
[/FI[LES]=documents]
[/OP[TION]= SOURCETARGETMATCH]
[/OUT=outfile]
[/OV[ERWRITE]=YES/NO]
[/QUIET]
Translation memory match information is counted either if the "Automatic substitution" option or the "Count Translation Memory match" option of analysis is selected. If the automatic substitution is used the exact matches in the document are replaced by data from the Translation Memory as if the segments had been translated manually; ie. the Translation Memory is updated too. Using the "Count Translation Memory matches" option the document and the Translation Memory remain unchanged. It is therefore possible to repeat the analysis several times without side effects.
In the list window containing the files of a folder, the Time and Date of the file can be displayed.
This "Updated (Time and Date)" and "Updated (Date)" should not be changed when the document is closed without update.
The "Time and Date" is only updated when the file is updated.
The enables the user to export a folder with or without Translation Memory databases defined on document level.
There is a new checkbox on the folder export dialog labeled "With document Translation Memory databases". The checkbox is only active if the "With Translation Memory databases" checkbox is selected. If the "With document TM" checkbox is selected all TMs defined on document level are exported together with the folder. If the checkbox is not selected only the TMs defined on folder level are exported with the folder.
In the batch API the new option is specified using the keyword "DOCMEM" in the options for the task "FOLEXP". If the option is not specified only the folder TMs will be exported.
The shows the number of terms organized during dictionary organize in the organize completion message similar to the number of segments organized during Translation Memory organize.
The completion message of the dictionary organize process has now one more parameter: The number of terms successfully organized.
The EQFTMT is a standalone program to change the markup and/or the target language of segments stored in a Translation Memory. Either a specific markup or target language can be changed to a new value or all segments regardless of the current settings can be modified.
The input parameter for the program EQFTMT are specified using the command line with the following syntax:
EQFTMT /MEM=memdb [/FROMMARKUP=|/FM=frommu] [/TOMARKUP=|TM=tomu]
[/FROMLANG=|/FL=fromlang] [/TOLANG=|/TL=tolang] [/TYPE=NOCONF]
where:
|
memdb |
is the name of the Translation Memory being processed |
|
frommu |
is the name of the markup which is to be changed. The name may contain wild card characters (e.g. /FROMMARKUP=IBM* would change all segments having a markup starting with IBM). If this parameter is omitted the markup of all segments is changed regardless of the current markup value. |
|
tomu |
is the new markup which is to be used for the segments having a matching markup or for all segments if the parameter /FROMMARKUP is omitted. |
|
Fromlang |
is the target language which is to be changed. The name may contain wild card characters (e.g. /FROMLANG=Eng* would change all segments having a target language starting with Eng). If this parameter is omitted the target language of all segments is changed regardless of the current target language value. |
|
tolang |
is the new target language for the segments having a matching target language or for all segments if the parameter /FROMLANG is omitted. |
If the paramater /TYPE=NOCONF is specified no processing window and not end message is shown otherwise the following process window is shown while the segments of the TM are changed:
In the process window the name of the Translation Memory, the number of the current segment, the number of segments with changed markup and the number of segments with changed target language is displayed.
At the end of the process a completion message is displayed showing the number of processed and changed segments.
During folder import the imported TMs are automatically merged if the TMs already exists. For dictionaries a "Merge confirmation message" is displayed. The will also display a merge confirmation message for imported TMs which already exist. The user can when select if the imported TM should be merged into the existing TM or if the imported TM should be ignored.
A new query message is displayed during folder import for each imported TM which already exists. The user can select "No" to ignore the imported TM or "Yes" to merge the imported TM and the existing TM. During batch folder import this message is not displayed and the merge is always performed.
This functions allows to print the contents of the currently active list window using the currently active view details.
The print list function is activated be selecting the menu item "Print list..." of the "File" pulldown while the input focus is on the list window which is to be printed.
The layout of the printout depends on the currently active list window and the selected view details. As an example here the printout for the following Translation Memory list window:
Printout:
Translation Memory List
Name Size Drive Source Language
ARCAL 40960 H: English(U.K.)
AXSYMF 12808192 H: English(U.S.)
AXSYMFX 36864 H: English(U.S.)AXSYMFY 36864 H: English(U.S.)
CODE37 15335424 H: English(U.S.)
CODE37N 15335424 H: English(U.S.)
CODE37Y 212992 H: English(U.S.)
DDS 1945600 H: German(national)
DEMO1 61440 H: English(U.S.)
IMXF 8835072 H: English(U.S.)
IMXFX 40960 H: English(U.S.)
KBT0116 2740224 H: English(U.S.)
KBT0119 36864 H: German(national)
NTCNT2 184320 H: English(U.S.)
RTF 36864 H: English(U.S.)
SAMPLE1 69632 H: English(U.S.)
TEST 90112 H: German(national)
TEST11 28672 H: English(U.S.)
TSTXPG4 36864 H: English(U.S.)
VOLVOTST 40960 H: English(U.S.)
The printout uses the same columns and column widths as the list window. If the list window has auto-text columns (e.g. for long document names) the width of the printed columns are also adjusted to the actual text in the columns.
If the user selects a file name in the files listbox of the import dialog, the name of the selected file (without file extension) presets the 'import into' field of the dialog. If the user does not want to import the file under the preselected name he has to overwrite the name in the 'import into' field.
For some specific column based formats it is important to see or position according to the cursor position. For this purpose a ruler similar to the one provided in EPM or LPEX is provided.
The ruler can be activated in the translation window, in the original window and for other documents" windows. The display is similar to the one provided in EPM or LPEX.
The ruler bar will be (and has to be) of the same size as the currently selected font.
The active cursor position is marked via a black block character (on gray background) allowing to easily determine the currently active cursor position.
In addition digits are displayed (at each deciaml position) to allow to position on the screen
The ruler can be switched on and off via an option in the context pop-up. For the translation window it is added to the TAILOR option, where a user can already decide if he wants to see a titlebar or/and the horizontal/vertical scrollbars.
For the original window or the window for other documents, the ruler is added to the context pop-up directly as can be seen below.
A translator’s notes function displays helpful commentary text to the translator. The notices are displayed on request in a forth window.In the editor window a special short tag indicates that a trnote exists at this position.
The implementation follows the specifications in the design, see therefore TRNOTDES.DOC:
The Translators notes window can be invoked
The keystroke is user-definable via the keys... dialog.
<TRNOTE> is visible in all displays styles: in hide mode, in shrink and compact mode.
<TRNOTE> is protected, and the coloring of it is the color selected for protected parts.
The user cannot overtype it, only in unprotected style.
When <TRNOTE> is deleted during translation of a segment, no message pops up that protected parts have been deleted.
The <TRNOTE> abbreviation is user-defineable in the profile settings dialog.
The translators notes window is sizeable and moveable.
If automatic linewrapping is on, the text in the notes window is also automatically wrapped.
The translators notes window always contains all notes which are contained in the current document. The translators notes window is available also via the Files / windows pulldown, once it has been invoked during an editing session.
If the user double-clicks on a special note in the translators notes window, the focus is set to the translation window and the segment, which is next to the specified translators note.
If a translators note is of level two, it is displayed in light grey ( default). For level one notes, black and blue are used alternatingly as foreground colors. The colors are user defineable in the colors dialog, as can be seen below.
The <TRNOTE> and the text which is the note, is not saved in the translation memory. Therefore a proposal and the source of proposals can never contain a <TRNOTE>. Also in the source of the current segment which is displayed in the proposal window, the translators notes have been deleted. This has the impact, that ctlr-0 copies the current source, but without the translators notes. Nevertheless the translators notes window always contains all notes which are contained in the document.
Usually the user wants to know where the cursor is positioned in the document. This is needed specially when the user wants to discuss with another translator about a special position.
The current approach, that the line can be asked for in the "Query line" dialog, is too cumbersome.
Therefore it is necessary, that the current line, column and segment is displayed in a status bar.
The current cursor position will be shown in a status line of the edit window. The status line can be switched on and off via an option in the context pop-up. It is added to the TAILOR option, where a user can already decide if he wants to see a titlebar or/and the horizontal/vertical scrollbars.
In addition to the line and column position the status bar will contain the segment number and the status of the cursor (Insert/Replace).
To have a static display of the statusbar we’ve chosen the approach, to reserve already some space, so that the size of the boxes will not change quite often. They will have space for 3 digits for the segment and the line number and 2 digits for the column position. The numbers are prefixed with 0s if they do not fill up the space. If any of the numbers exceed the reserved space, the field is dynamically adjusted to contain the whole number.
For determining the cursor status, we display one status (Ins) and grey it, if the user switches to overlay mode.
The fields are right adjusted, the rest of the status bar is left free and may be later used to display status information (as it is done in some other tools).
The status bar is attached to the Translation processor (Standard Editor) window(s). Another approach was thought through (Attaching the status bar to the Translation Workbench), but not chosen, because of implications with other attached editors, like e.g. XLATE.
This displays the entire document within the window size without actually inserting character in the document. The document is wrapped according to the window size, and the wrapping is actualized at once if the editor window is resized.
Thus all text is readable without scrolling the editor window to the right.
This can be easily seen in the following two bitmaps where the same text is wrapped according to different window size.
The new wrapping will insert "Soft linebreaks". These will not be stored in the exported file, in the segmented file, or in the Translation Memory databases.
In the profile settings dialog, the user can now select a right margin which is a number ( 50, 60, ...) or AUTO.
If the right margin AUTO is selected, and "Linewrap" is ON in the edit pulldown, soft linefeeds are inserted in the entire document.
If the right margin is any number, and "Linewrap" is ON in the edit pulldown, the editor behaviour is as usual. Not the whole document is wrapped, only the current segment is wrapped during typing.
A further new feature of the profile settings dialog is to select the style of the Translation Memory window.
In PostEdit mode it is not possible to see where the segment ends, specially if more than one segment are in one line.
This new enhancement displays a ¶ at the end of a segment, if the user is in PostEdit mode and has requested this segment boundary sign in the settings dialog.
The segment boundary sign is defined to be ¶.
This sign is not user-definable.
It is protected in any style, which means it cannot be overwritten in UNPROTECT style.
Hence the display is the same color as for protected characters.
This sign is added if the user is in PostEdit mode and has selected the option „Segment boundary sign" in the settings dialog. If the user activates a segment or turns off the specified option in the settings dialog, all signs are removed at once.
The boundary sign is added at the end of each segment as can be seen in the following bitmap.
This option will allow you to actively work on more than one document the same time in the Translation Environment.
The user can select multiple documents in the document list or open additional files to be translated via the Open Files dialog from within the Standard editor
The associated Translation Memory will be updated if you change/translate a sentence.
The user can select multiple documents in the document list (as long as they are using the same editor). Opening the Translation Environment will open the first one of the selected documents. The user now can open the additional ones if they select the new Translation Environment option in the file pulldown of the Standard Editor and select the appropriate document. This approach was chosen (instead of loading all documents at once) to avoid long loading times, overloaded Translation Environment with a lot of windows and memory restrictions and the limited list of file handles.
Toggling between the different active Translation windows is achieved either via clicking on the appropriate Translation window or via selecting the document from within the Translation Environment option of the File pulldown.
After toggling the dictionary/dictionaries as well as the Translation Memory of the newly active document is activated and the appropriate sentence is looked up in the Translation Memory and dictionary.
In addition the user can select (at any time) additional documents (either via selecting documents from a document list or via opening a document with the File Open dialog from the Standard Editor).
Closing a Translation window will close the target and the source file. If there is still one file in the list of to be opened documents, the next one in the list will be opened. If not, the Translation Environment will be closed.
Usually the user wants to search for a word in the whole document, not only from the current position until the bottom or top of the document.
Therefore the Find&Change dialog should not stop at the end of the document but wrap around to the begin of the document until the position is reached form which the Find&Changefunction has been activated.
Also the backward search should not stop at the beginning of the document but wrap around to the end of the document and continue the search until the start position is reached.
This makes the searching through the whole file much more easier.
The Find & Change function consists internally of two functions, one for forward searching and one for backward searching. This has been necessary for the DBCS enabling of the search function.
Therefore the forward searching routine has to be adapted and the backward routine has to be adapted separately.
To know the starting position, the starting segment number and the starting segment offset are remembered in two new fields in the dialog structure "FINDDATA".
When "Change" is activated by the user, the starting position must be updated if the change happened in the starting segment in front of starting segment offset.
The Find & Change dialog is a modeless dialog.
However the starting segment number and segment offset are not changed, as long as the Find & Change dialog is open, but no changes in the dialog settings occurred.
The starting segment number and offset are updated when the Find / Change term is changed.
The change range in the Find & Change dialog now has the option "File" instead of "Current to End".
The users need a function which allows to position to the next irregular ( i.e. non 1:1 ) match. An irregular match is a match which is either
- 1:2, 2:1 or 2:2 match ( no special color)
- unaligned sentence ( default color red)
- an ignored ( i.e. crossed-out) sentence ( default color grey).
This gives the user the possibility to check all alignments where not exactly one sentence has been aligned with one other sentence.
In the " cursor" pulldown, the " Next irregular" and " Previous irregular" is added.
For " Next irregular" the keystroke " Shift-Alt-Down" is determined, for " Previous irregular" the keystroke " Shift-Alt-Up" is determined.
After the user selected " Next irregular", the program searches it, positions at the alignment next to it, and marks the " irregular" segment. Marking the irregular sentence, is an increase in performance, because the user can now realign this segment at once. ( All features in the " Edit" pulldown require that the corresponding sentences are marked.)
If there are several segments following each other, which are all ignored segments and/or unaligned segments, only the first one is marked and positioned to. The next „Next irregular" positions to the next irregular segment after the continuous heap of unaligned or ignored segments.
If the search is at the end of the document, no wrap-around is done.
This does not position at " over-crossed" alignments. This is not necessary because these can only be generated by the user and occur very, very rarely. Therefore these over-cross alignments are already verified and controlled by the user.
The new profile settings dialog replaces the old profile settings dialog as well as the right margins dialog and the profile initials dialog.
Therefore the user can make all user settings in the profile settings dialog and all settings are active at once.
This simplifies the current approach, where the user definable settings are spread over several dialogs.
The ifollowing items have been added to the " Profile settings" dialog:
- Line number for active segment
- Right margin
- Translation Memory window style
All settings in the profile settings dialog are active at once, i.e. when the user has selected „Set" in the dialog.
In the " Right margin" entryfield, the user can now select a right margin which is a number ( 50, 60, ...) or AUTO. Also it is possible to input any number in the valid range for a right margin.
The items in the listbox of the profile settings dialog have been rearranged. First are the functions which take place in the editor, then the functions modifying the proposal in the Translation Memory window, then the functions dealing with automatic substitution, and at the end functions for the dictionary window.
When TranslationManager starts again, always the last-used values of the user settings are active.
The Translation Memory property dialog displays the properties of a Translation Memory and allows to modify the Translation Memory description.
Invocation
The Translation Memory dialog is activated either selection "Properties..." of the "File" pulldown while a Translation Memory in the Translation Memory list window is selected and the Translation Memory list window has the focus
or by selecting "Properties..." of the Translation Memory list window popup menu
In both cases the Translation Memory property dialog window is displayed.
Translation Memory Property Dialog
The dialog window displays the following information for the selected Translation Memory:
|
Name |
The name of the Translation Memory |
|
Location |
"Local" for local TMs or "Shared" for shared TMs |
|
Disk drive |
The drive of the TM database files |
|
Source language |
The TM source language |
|
Segments |
The total number of segments stored in the TM |
|
File size |
The overall size of the TM database files |
|
Last update |
The date and time of the last update of the TM |
|
Documents |
A list of all documents for which segments are stored in this TM |
|
Markups |
A list of markups which are used by the segments of this TM |
Only the description field may be modified by the user. Selecting "Change" saves any modified description and leaves the dialog. Selecting "Cancel" ends the dialog without saving any changes.
The base TM/2 product is now included in one dll instead of approx 20 dll's for TP4.7 and lower. This results in higher stability in terms of packaging and installation and in faster response times.
All user exists are still included in single dll's.
The following dlls are obsolet and will be deleted by the install processThe rename function allows to rename Translation Memory databases, dictionaries and folders and to adjust references to the renamed object automatically
The Translation Memory match counting function allows to count the number and type of matches which can be achieved using a specific Translation Memory on a specific document, a group of documents or all documents of a folder. Unlike "Automatic substitution" the does not alter the document(s) nor the Translation Memory and supplies also information about fuzzy matches.
There are two steps required in order to get the Translation Memory match information:
For the Translation Memory match information either the "Substitute exact matches in the document automatically" option or the "Count Translation Memory match information" option must be selected.
EQFCMD /TA[SK]=ANALYZE
/FLD=folder
[/FI[LES]=files]
[OP[TION]=( [ADDTOMEM,]
[AUTO,]
[UNTRANSLATED,]
[AUTOLAST,]
[AUTOJOIN,]
[AUTOCONTEXT,]
[TMMATCH])
[/MEM=mem]
[/OV[ERWRITE]=[YES|NO]]
[/QUIET]
For Translation Memory match information either the "AUTO" or the "TMMATCH" option must be selected.
Word Count step using the GUI
The option "Translation Memory match information" has to be selected to display the Translation Memory match data.
In order to count replace matches and fuzzy replace matches separately the option "Count replace matches separately" has to be selected.
Result of Word Count using the GUI
The following information is displayed in the window:
Column 1: The name of the document
Column 2: The total number of words in the document
Column 3: The number of words in segments with exact-exact matches
Column 4: The number of words in segments having exactly one proposal
Column 5: The number of words in segments with more than one exact proposal
Column 6: The number of words in segments with fuzzy proposals
Column 7: The number of words in segments having no proposal at all
If the option "Count replace matches separately" has been selected two additional columns are displayed:
Column 8 The number of words in segments with replace matches
Column 9 The number of words in segments with fuzzy replace matches
If this option has not been set the number of words in replace matches is contained in the number of words in the fuzzy match column.
All columns are "autowidth" columns; i.e. the columns are automatically resized to display all information without truncation.
If there is no Translation Memory match data available for a document (either because the document has not been analyzed yet or because neither the "Automatic substitution" nor the "Count TM Match Info" option has been selected for the analysis) the output line for the document contains the words "No match info" rather than the word count results.
Any changes in the Style pulldown will be reflected directly in
the Translation Proposal as well as in the default value shown
in the Profile Settings dialog.
Will be included in next version !
!
The table of contents is displayed in a separate listbox. Double-clicking on an item in the table of contents positions the translation window on this text.
When the user selects an entry in the table of contents and presses the Goto pushbutton or double-clicks on the listbox-item the cursor is positioned on the spedified text in the document.
The members of the table of contents can be provided by a new user-exit function EQFTOCGOTO.
The user-exit is called when the user invokes the "Special goto..." or the option "Special goto dialog" is selected in the commands dialog or via key setting specified in the keys dialog.
Helper functions like EQFGETNEXTSEG,EQFGETPREVSEG, EQFGETCURSEG and EQFADDTOTOC are provided so that the user exit can retrieve the text of the whole document. Then it is up to the user exit to retrieve and check the text which should be displayed in the table of contents. This text, combined with a reference to the containing segment number, must be send back to the TranslationManager using the helper function EQFADDTOTOC .
Then TranslationManager displays the table of contents.
If no user-exit function is available, TranslationManager loops through the document and displays all entries which are marked in the markup table with the new classid cls_head.
If no user-exit function is available, TranslationManager loops through the document and displays all entries which are marked in the markup table with the new classid cls_head. An example how the table of contents is filled using this classid, has been provided for IBMIDDOC. Therefore the classid "cls_head" has been added to the following tags:
<TAG>
<STRING><BIBENTRY</STRING>
<ENDDELIM>' >\r\n'</ENDDELIM>
<TYPE>STNEUTRAL</TYPE>
<ATTRINFO>YES</ATTRINFO>
<CLASSID>CLS_HEAD</CLASSID>
</TAG>
<TAG>
<STRING><DOCTITLE</STRING>
<ENDDELIM> >\r\n'</ENDDELIM>
<TYPE>STNEUTRAL</TYPE>
<ATTRINFO>YES</ATTRINFO>
<CLASSID>CLS_HEAD</CLASSID>
</TAG>
<TAG>
<STRING><IBMBIBENTRY</STRING>
<ENDDELIM>' >\r\n'</ENDDELIM>
<TYPE>STNEUTRAL</TYPE>
<ATTRINFO>YES</ATTRINFO>
<CLASSID>CLS_HEAD</CLASSID>
</TAG>
<TAG>
<STRING><TITLE</STRING>
<ENDDELIM>' >\r\n'</ENDDELIM>
<TYPE>STNEUTRAL</TYPE>
<ATTRINFO>YES</ATTRINFO>
<CLASSID>CLS_HEAD</CLASSID>
</TAG>
<TAG>
<STRING><TITLEBLK</STRING>
<ENDDELIM>' >\r\n'</ENDDELIM>
<TYPE>STNEUTRAL</TYPE>
<ATTRINFO>YES</ATTRINFO>
<CLASSID>CLS_HEAD</CLASSID>
</TAG>
The API entry point function EQFTOCGOTO must be provided by the user exit to support the display of a table of contents.
The function EQFTOCGOTO is invoked if the user selectes "Special goto..." within the editor.
Format
| sResult = EQFTOCGOTO ( lInfo, hwndLB ) |
Where
lInfo(LONG)- input
the information which must be given back to the helper function EQFTOCGOTO as first parameter
hwndLB(HWND)- input
handle of listbox where data has to be added; handle must be forth parameter in EQFTOCGOTO
Purpose
This function must retrieve, prepare and send back the data for the table of contents using the helper functions.
Return Codes
sResult - number of entries in table of contents
An example how the user-exit can fill the table of contents, has been provided for Bookmaster.
Format
| usResult = EQFADDTOTOC ( lInfo, pusSegNum, pchData, hwndLB ) |
Where
lInfo(LONG)- input
the information which is contained in the first parameter of EQFTOCGOTO
pusSegNum(PUSHORT)- input
segment number of the segment to which the data belongs. Double-click on the data (sent in pchData) in the table of contents causes the cursor to be positioned on this segment.
pchData(PCHAR)- input
data added as next item into table of contents list. This buffer should not exceed the length of 128 characters.
hwndLB(HWND)- input
handle where data has to be added; handle is known as 2nd parameter of EQFTOCGOTO
Purpose
This call will fill the next item into the table of contents list.
Return Codes
0 - no error occurred
Remarks The first and the last parameter are known as first and second input parameter of a previous EQFTOCGOTO call. Therefore this function can only be used to fill the table of contents.
The buffer which contains the data, must be allocated and freed by the user-exit.
The other helper functions EQFGETNEXTSEG, EQFGETPREVSEG and EQFGETCURSEG are described in the document for the WYSIWYG support.
If the first protected word is longer than 10 characters, the word is cut off after 10 characters.
The following 3 screenshots show a document with markup IBMIDDOC in Protected style, in Compact+1 style and in Compact style.
In protected style, some tagging is very long.
In compact+1 style, the tagging is short, but the user can still guess the meaning of the tag.
In compact style, the tagging is very short, but the user has no idea what kind of tagging it is.
